Boada, R. 1, 2 , Sánchez-Casas, R. 1, 2 , Ferré, P. 1, 2 , Guasch, M. 1, 2 & García-Albea, J. E. 1, 2
1 Departament de Psicologia. Universitat Rovira i Virgili. Tarragona, Spain.
2 CRAMC. Tarragona, Spain.
Word ambiguity across languages (i.e., multiple translations) has been found to slow down response times and to reduce accuracy in translation tasks (Boada et al., 2010; Laxén & Lavaur, 2010; Sánchez-Casas et al., 1992; Tokowicz & Kroll, 2007). Multiple translation words usually have a dominant translation (the one mainly given by participants when asked) and many subordinate translations. Previous findings on translation dominance using a translation recognition task found that participants were more error-prone when recognizing subordinate translations, but the results differ across studies regarding response times. Moreover, there is recent evidence that suggests the semantic relation between dominant and subordinate translations can influence the translation performance (Laxén & Lavaur, 2010).
The aim of the present study was to further examine how translation dominance and semantic relatedness between the possible translations of a word affected bilingual word recognition. Two sets of Catalan and Spanish words with multiple translations into the other language were selected. For a given word, translation dominance was established on the basis of three measures obtained through different questionnaires. These measures provided information regarding: a) possible word translations; b) translation recognition; and c) an estimation of how frequent each translation was used. Additionally, ratings concerning the semantic relatedness between the dominant and the subordinate translations were collected.
The multiple translation words from the two sets (Spanish-Catalan and Catalan-Spanish translation pairs) were used as targets in a lexical decision task. They were presented to the same group of highly proficient bilinguals, using a priming paradigm so that each word could be preceded by both their dominant and subordinate translations. The relevant analyses compared priming effects as a function of translation dominance and semantic relatedness between the translation word pairs. The results are discussed within the framework of the Distributed Conceptual Representation Model (de Groot, 1992).